INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

zanne

-0.79

eters

-0.71

obyl

-0.71

inyl

-0.69

amin

-0.65

 embr

-0.63

forts

-0.63

romy

-0.63

amina

-0.63

ensitive

-0.63

POSITIVE LOGITS

TeX

0.74

Brow

0.74

Tumblr

0.69

neau

0.69

Ń·

0.69

onian

0.66

SON

0.64

 comma

0.62

 Spice

0.61

 conquest

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.